Archives of Biochemistry and Biophysics 496 (2010) 1–8

Contents lists available at ScienceDirect

Archives of Biochemistry and Biophysics

journal homepage: www.elsevier .com/ locate/yabbi

Sites in the AAV5 capsid tolerant to deletions and tandem duplications

Kaoru Hida a, Sang Y. Won b, Giovanni Di Pasquale c, Justin Hanes b,d, John A. Chiorini c, Marc Ostermeier b,d,*

a Department of Biomedical Engineering, The Johns Hopkins University, Baltimore, MD 21218, USAb Department of Chemical and Biomolecular Engineering, The Johns Hopkins University, Baltimore, MD 21218, USAc Gene Therapy and Therapeutics Branch, National Institute of Dental and Craniofacial Research, National Institutes of Health, Bethesda, MD 20892, USAd Institute for NanoBioTechnology, The Johns Hopkins University, Baltimore, MD 21218, USA

a r t i c l e i n f o

Article history:Received 3 November 2009and in revised form 18 January 2010Available online 25 January 2010

Keywords:Directed evolutionAAVViral gene delivery

0003-9861/$ - see front matter � 2010 Elsevier Inc. Adoi:10.1016/j.abb.2010.01.009

* Corresponding author. Address: Chemical and BioHopkins University, 3400 N. Charles Street, 221 MDUSA. Fax: +1 410 516 5510.

E-mail address: oster@jhu.edu (M. Ostermeier).1 Abbreviations used: AAV, adeno-associated virus;

Eagle medium; FBS, fetal bovine serum; rAAV2, recomb

a b s t r a c t

Gene therapy vectors based on adeno-associated virus (AAV) have shown much promise in clinical trialsfor the treatment of a variety of diseases. However, the ability to manipulate and engineer the viral sur-face for enhanced efficiency is necessary to overcome such barriers as pre-existing immunity and trans-duction of non-target cells that currently limit AAV applications. Although single amino acid changes andpeptide insertions at select sites have been explored previously, the tolerance of AAV to small deletionsand tandem duplications of sequence has not been globally addressed. Here, we have generated a large,diverse library of >105 members containing deletions and tandem duplications throughout the viral cap-sid of AAV5. Four unique mutants were identified that maintain the ability to form viral particles, withone showing improved transduction on both 293T and BEAS-2B cells. This approach may find potentialuse for the generation of novel variants with improved and altered properties or in the identificationof sites that are tolerant to insertions of targeting ligands.

� 2010 Elsevier Inc. All rights reserved.

1. Introduction

Adeno-associated virus (AAV)1 is considered one of the mostpromising vectors for gene therapy due to its non-pathogenic,non-toxic nature and its ability to achieve long term gene expres-sion in a variety of cell lines, tissues and animal models [1–5].However, despite early promise in clinical trials with recombinantAAV serotype 2 (rAAV2) vectors [6,7], there still remain a numberof barriers that limit the efficacy of AAV vectors. For example, theheparin sulfates, which are the primary receptors for AAV2, arewidespread, leading to difficulties in achieving good tissue disper-sion or specific cell targeting [8]. Alternatively many therapeutictarget cells express very low levels of receptors, necessitating highvector doses. The apical surfaces of airway epithelial cells, a com-mon target for many lung diseases, typically have limited expres-sion of viral receptors [9–11]. Finally, the pre-existence ofneutralizing antibodies against AAV in the general population,�32% of individuals in one study [12], may limit initial efficacyand hinder repeated administrations. Thus, there is a need to beable to engineer viruses to specific target cells, enhance infection,

ll rights reserved.

molecular Engineering, JohnsHall, Baltimore, MD 21218,

DMEM, Dulbecco’s modifiedinant AAV serotype 2 vectors.

reduce non-specific transduction, and overcome pre-existing neu-tralizing antibodies and other extracellular barriers.

Earlier attempts to understand and engineer the AAV surfacehave predominantly focused on serotype 2. Mutagenesis studieshave identified sites amenable to insertion of foreign peptides[13–16] as well as sites critical for AAV2 receptor [16,17] and anti-body binding [18–20]. Single amino acid changes on the AAV sur-face can dramatically change such phenotypes as viral titer,receptor binding, and tissue tropism exhibited by closely relatedAAV serotypes [21]. These early studies often involved the individ-ual construction of many capsid mutants and subsequent identifi-cation of new phenotypes. For example, 38 virus capsid mutantscontaining peptide insertions at 25 unique sites within the AAV2capsid were constructed by Shi et al. to identify insertion sites[14] while Wu et al. [16] constructed 93 mutants at 59 differentpositions in the AAV2 capsid gene by site directed mutations. Morerecently, directed evolution approaches have been utilized to intro-duce point mutations randomly throughout the capsid and subse-quently selected for enhanced properties such as neutralizingantibody evasion [22,23]. However, the tolerance of AAV to ran-domly introduced deletions or insertions has not been addressed.

Serotypes other than AAV2 have been shown to have specificand superior transduction efficiencies for a variety of tissues [24–26]. For example, AAV5 may be the vector of choice for gene ther-apy targeted for the lungs; it transduces the lungs more efficientlythan serotype 2 [27,28] and recognizes PDGFR [29], a receptorexpressed on the apical surface of the airways. AAV5 is the most

2 K. Hida et al. / Archives of Biochemistry and Biophysics 496 (2010) 1–8

distant serotype within the parvovirus family, with only a 54–56%sequence homology to AAV2 [30,31], but like all the other AAV ser-otypes, it is composed of three structural proteins designated VP1,VP2, and VP3 present at a ratio of 1:1:10. These capsid genes are allencoded in the same open reading frame and expressed from thesame viral p40 promoter. The amino acid sequence of the majorcapsid protein VP3, contained within the two larger capsid pro-teins, is the primary protein that constitutes the surface topologyof the virus. Here we have randomly introduced deletions and tan-dem duplications (rather than point mutations) on the AAV5 cap-sid to examine how tolerant AAV5 is to such mutations and toidentify sites on the capsid that may be amenable to the insertionof targeting peptides.

2. Materials and Methods

2.1. Cell culture

Low passage 293T cells were maintained in Dulbecco’s modifiedEagle medium (DMEM) supplemented with 10% FBS and 1% peni-cillin/streptomycin (Pen/Strep). BEAS-2B cells were maintained inDMEM/F12 supplemented with 10% fetal bovine serum (FBS) and1% Pen/Strep. All cells were cultured at 37 �C in a 5% CO2 environ-ment. For screening, cells were seeded at a density of 3.5 � 106

cells in 15 cm plates and for flow cytometry studies cells wereseeded at 10,000 cells/well in 24-well plates and allowed to growfor 24 h before manipulation.

2.2. Plasmids

The pDIM-N2 plasmid [32] served as the backbone plasmid formost of the library manipulation. SpeI and MluI restriction enzymesites were introduced outside of the BstEII and RsrII sites withinthe AAV5 cap gene by PCR on pACYC184-AAV5 plasmid which con-tains the full sequence of wild-type AAV5 (1–4642 nt). This fragmentwas subsequently moved into the pDIM-N2 plasmid to yield pDIM-N2-AAV5. The BtgzI restriction enzyme site within the AAV5 capgene was silently removed using the QuikChange Site-DirectedMutageneis Kit (Stratagene) per manufacturer instructions usingthe primers, forward 50-CACCATAGCCAACAACCTCACCTCC-30 andreverse 50-TGTTGGCTATGGTGGTGGTGGAGTC-30.

A plasmid containing the ‘marker’ sequence was created suchthat PvuII digestion would yield the desired marker fragment shownin Fig. 1B. The marker sequence was derived from a fragment of thegene encoding YFP. Primers were used such that a PvuII restrictionenzyme site would start 7 bp away from the end of the BtgzI site(forward 50-TCATCGCAGCTGCAGCTTACATCGCCCACCTACGGCAAGCTG-30 and reverse 50-TCATCGCAGCTGCAGCTTACATCGCCTTGTACAGCTCGTCCATG-30). The primers were used to amplify a portionof the YFP gene, and transferred into the pDIM-N2 plasmid. Afterthe plasmid was transformed into DH5a cells, amplified and puri-fied, digestion with PvuII yielded the blunt ended ‘marker’ sequence.

For the final step of the library, all mutations made on the pDIM-N2-AAV5 plasmid needed to be transferred back into the pACYC184-AAV5 plasmid with the rest of the cap and rep genes. To minimizeany transfer of un-digested wild-type capsids into the final library,an internal stop codon and unique restriction enzyme site (AflII)were introduced into the cap gene, following the instructions onthe Phusion site-directed mutagenesis kit (New England BioLabs)using the primers, forward 50-CAAAATCTTCAACACTTAAGTCAAAGAG-30, and reverse 50-ACTCTGAGGGACCGGGGTCTG-30 primers.The library in the pDIM-N2-AAV5 plasmid was transferred into thismodified pACYC184-AAV5 plasmid instead of the wild-type. Thisfinal library was digested with AflII to further ensure there were

no wild-type plasmids arising from incomplete digestions beforethe transfection/screening steps.

2.3. Construction of the plasmid library

For the initial step in the library process, the pDIM-N2-AAV5plasmid was treated with serial dilutions of diluted DNase I (1.6–0.0016 mU/ug DNA) in a buffer containing 1 mM MnCl2, 0.05 MTris–HCl and 0.05 mg/mL BSA. The reaction was allowed to takeplace at 22 �C for 10 min until the reaction was stopped by theaddition of 1.2 lL of 1 M EDTA to 100 ll of the reaction mixtureand heating to 75 �C for 10 min. The reactions were analyzed byagarose gel electrophoresis to determine the concentration ofDNase I that yielded the largest amount of linear fragments butalso with a fair amount of supercoiled DNA, indicating that theDNA had not been over-digested. After gel purification (Qiagen)of the linear fragment, the ends were then repaired with 1 U T4DNA polymerase and 160 U T4 DNA ligase/lg of DNA in ligase buf-fer with 200 lM dNTPs. The reaction proceeded for 20 min at 12 �Cuntil it was terminated by the addition of EDTA to a final concen-tration of 10 mM. The ‘marker’ sequence was ligated into the line-arized and repaired library with T4 DNA ligase overnight at 16 �C.Ligations were done in 10 lL reactions with 2000 U T4 DNA ligase,100 ng of the repaired library and 30 ng of the ‘marker’ sequence(�1:3 mole ratio) in T4 DNA ligase buffer and 5% w/v of 8000 MWPEG. The plasmids were amplified by transformation into electro-competent DH5a Escherichia coli cells and the resulting plasmid li-brary purified (Qiagen). This initial library containing the ‘marker’ se-quence was then digested with BstEII, RsrII, AseI and AhdI. The largestfragment was gel purified (Qiagen) and ligated into another pDIM-N2plasmid in the same fashion as described above. Another transforma-tion amplification and purification yielded a plasmid library contain-ing the AAV5 capsid with the ‘marker’ sequence contained within.This ‘marker’ sequence was removed by enzymatic digestion withBtgzI. The linear fragment was gel purified and repaired as describedabove. Overnight ligations were done in 10 ll reactions with 100 ngof the repaired library to recircularize the plasmid.

2.4. AAV5 library Screening/production and titration

For the production and screening of the AAV5 library, 293T cellsin a 15 cm diameter plate were transfected, by Polyfect TransfectionReagent (Qiagen) according to manufacturer instructions, with13 lg of an adenovirus helper plasmid (pAd12) containing the VARNA, E2, and E4, and 6, 60, 600 ng or 6 lg of the plasmid library.The DNA was pre-mixed and added to polyfect reagent at a ratio of10 ll:1 lg. After 48 h, the cells were harvested, lysed via freeze/thaw cycles and addition of 0.5% w/w deoxycholic acid, and free nu-cleic acids were digested with benzonase. After the cell debris wasremoved by centrifugation, the recovered virus was purified usinga CsCl gradient. The nuclease resistant genomic particle titer wasdetermined by quantitative real-time PCR using the TaqMan system(BioRad) with probes specific to the AAV5 wt genome (forward50-ACCAACATCGCGGAGGC-30 and reverse 50-CCCTCCTCCCACCAAATGA-30). Viruses in CsCl were dialyzed 24 h using 0.5 ml Slide-A-Lyzer(Pierce) in 500 ml of PBS with changing of the buffer three or fourtimes.

2.5. Sequencing of AAV library clones

The dialyzed viral library was collected and the AAV served as atemplate for PCR. PCR products were analyzed by gel electrophoresis,digested with RsrII/BstEII, cloned into Topo- AAV5 cap (AAV5 nucle-otides 1700–4500 nt) which contains cap ORF and its endogenouspromoter. After transformation into E. coli, random colonies were se-lected for sequencing. All sequencing was done at Genewiz, Inc.

Fig. 1. Construction of the deletion/duplication library. (A) A plasmid containing the adeno-associated virus serotype 5 (AAV5) capsid gene between unique restrictionenzyme sites BstEII and RsrII was digested with dilute concentrations of DNaseI to create on average, one random double-stranded break per plasmid molecule. Afterpurification of DNA with a single double strand cut, ends were repaired to form blunt ends. Repair of 50 overhangs results in tandem duplications and repair of 30 overhangsresults in deletions. Recircularization of the vector occurred such that a ‘marker’ sequence was inserted at the site of the single double-stranded break. Digestion with BstEIIand RsrII yields two different fragment sizes containing the AAV5 gene, the larger one contained the ‘marker’ sequence. Purification of such fragments significantly decreasesthe number of wild-type sequences in the final library. The ‘marker’ sequence was subsequently removed using BtgzI digestion, repaired and ligated to create a library withrandom deletions and tandem duplications. (B) The ‘marker’ sequence contains two BtgZI restriction enzyme sites ten nucleotides away from either end. The enzyme cuts onestrand at the end of the marker and the other strand four nucleotides outside of the marker sequence, as indicated by the dark triangles. Upon digestion and repair by apolymerase to form blunt ends, the marker sequence is completely removed without leaving any non-AAV5 sequences behind.

K. Hida et al. / Archives of Biochemistry and Biophysics 496 (2010) 1–8 3

2.6. Recombinant AAV5b�actin-GFP production and titration

Recombinant (r) AAV5, and library clone viruses expressingb-actin-GFP were produced using a four-plasmid procedure as pre-viously described [33]. Briefly, semiconfluent 293T cells weretransfected by calcium phosphate with four plasmids: an adenovi-rus helper plasmid (pAd12) containing the VA RNA and coding forthe E2 and E4 proteins; two AAV helper plasmids containing theAAV2 rep and Topo-AAV5 with wt or library cap, respectively,and a vector plasmid containing AAV2 inverted terminal repeatsflanking an CMV b-actin-GFP expression cassette. Forty-eighthours post-transfection the cells were harvested by scraping intoTD buffer (140 mM NaCl, 5 mM KCl, 0.7 mM K2HPO4, 25 mMTris–HCl {pH 7.4}) and the cell pellet was collected by low-speedcentrifugation. Cells were lysed in TD buffer by three cycles offreeze–thaw. The clarified lysate (obtained by further low-speedcentrifugation) was treated with 0.5% w/w deoxycolic acid and100 U/ml Benzonase for 30 min at 37 �C. The virus was then puri-fied using CsCl gradients. Particle titers were determined by qPCR.Amplification was detected using an ABI 7700 sequence detector(ABI). Specific primers for CMV were designed by using the PrimerExpress program (ABI): CMV forward 50-CATCTACGTATTAGTCATCGCTATTACCAT-30, CMV reverse 50-TGGAAATCCCCGTGAGTCA-30.Following denaturation at 96 �C for 10 min, cycling conditionswere 96 �C for 15 s, 60 �C for 1 min for 40 cycles. The viral DNAin each sample was quantified by comparing the fluorescenceprofiles with a set of DNA standards. The rAAV particle titers were1–5 � 1012 DNAse resistant particles per ml.

2.7. Gene transfer assay in vitro

293T and BEAS-2B were incubated with purified wt or libraryclones of AAV5 b-actin-GFP at an MOI (genomic) of 3000. Fourdays later, cells were harvested with trypsin/EDTA and re-sus-pended in PBS. Reporter gene activity (5000–10,000 cells persample) was detected using FACSCaliber flow cytometer (BectonDickinson).

3. Results

3.1. Construction of the Library

To create our library of random tandem deletions and duplica-tions, we targeted the region within the AAV5 genome betweentwo unique restriction enzyme sites, BstEII and RsrII, which codesfor amino acids 156–666 (VP1 protein numbering). The overall li-brary construction scheme is depicted in Fig. 1A. The DNA frag-ment between BstEII and RsrII was initially transferred into asmaller plasmid, which was subsequently treated with diluteDNase I to generate, on average, one random double strand breakper plasmid molecule (steps 1 and 2). Double strand breaks createdby DNase I can have blunt ends, 50 overhangs or 30 overhangs. T4DNA polymerase, which has 50 ? 30 polymerase and 30 ? 50 exonu-clease activity, was used to make both types of overhangs blunt(step 3), resulting in tandem duplication of sequences from 50 over-hangs and deletions from 30 overhangs.

4 K. Hida et al. / Archives of Biochemistry and Biophysics 496 (2010) 1–8

Recircularization of the plasmid at this juncture would result ina library containing deletions and tandem duplications in the cap-sid gene fragment. However, since DNase I does not specificallytarget the capsid gene fragment, a large fraction of our librarywould contain deletions and duplications elsewhere in the plasmidand an unaltered capsid gene. This prevalence of wild-type capsidin our library would hamper our ability to identify functional caps-ids with deletions and tandem duplications. We minimized thispossibility by designing a ‘marker’ sequence, which was used to se-lect for those library members in which DNase I digested withinour target sequence (Fig. 1B). This marker sequence was insertedinto the plasmid after the repair step during cyclization of the plas-mid (step 4, Fig. 1A). After subsequent digestion with BstEII andRsrII (step 5), capsid gene fragments with a marker sequence wereisolated by agarose gel electrophoresis based on size (�2150 bp)and ligated into a new plasmid (steps 6 and 7). The marker se-quence was excised (step 8) using the restriction enzyme BtgzI.The marker sequence was designed to contain BtgzI sites near bothends. BtgzI digests 10 and 14 nucleotides outside its recognitionsequence; thus, the marker sequence could be cleanly removedleaving only DNA originating from the capsid gene (Fig. 1B). Oncethe marker was excised, a final repair step made blunt ends andDNA was re-circularized (steps 9 and 10) to make our final plasmidlibrary, which was comprised of 1.4 � 105 transformants.

Nine naïve library members were sequenced to verify success-ful library construction and examine the variety and the lengthsof deletions and tandem duplications (Fig. 2). The mutations ap-peared randomly distributed throughout the targeted gene(Fig. 2A) and both deletions and tandem duplications were found(Fig. 2B). More deletions than duplications were identified, withthe largest deletion being 119 bp and all others being much smal-ler, averaging about 18 bp. This bias towards deletions is typical ofrandom digestion libraries made with DNase I [34]. Since wild-typesequences would result when DNAse I digestion produced bluntends, we expected wild-type sequences to exist in the library.However, the fact that no wild-type sequences were foundamongst the nine naïve library members indicates that they arenot overly prevalent in the library and that our strategy tominimize wild-type sequences using the ‘marker’ sequence was

Fig. 2. Naïve Library Members. (A) Mutations were distributed randomly throughout theidentified. The numbers indicate the number of the deleted nucleotides (nucleotide 1 is thin the library are shown here. Clone M1 has a 19 bp deletion and clone M2 a 24 bp dup

successful. The majority of naïve library members contain frameshift mutations resulting in non-functional capsid proteins.

3.2. Screening for mutants able to assemble AAV5 particles

We evaluated our library on 293T cells to identify modifiedcapsids that were capable of assembling into a viral particle. Weexpected to isolate wild-type sequences despite our efforts to elim-inate them from the naïve library. Based on previous work on li-braries of this type [34], we expected that approximately 1–10%of naïve library members will contain neither deletions nor dupli-cations (i.e. they will be wild-type). The presence of non-mutatedsequences is unavoidable in many types of mutagenesis libraries.We were unsure whether functional mutant library memberscould be identified under this selection scheme, in which wild-typesequences will be strongly selected—both because wild-type in-fects these cells efficiently and because of the prevalence of wild-type sequences in the library. Functional mutants would have toappear with a significant frequency for our selection process to iso-late them.

The library was first subcloned into the original plasmid (pA-CYC184-AAV5 plasmid) with the remainder of the cap and thewild-type rep genes. 293T cells were then co-transfected withvarying concentrations of the library plasmid (6–6000 ng) and ade-novirus helper plasmid to determine the lowest concentration of li-brary plasmid needed for detectable viral production. Minimizinglibrary plasmid DNA during transfection helped ensure that theviral particles produced contained the DNA that coded for theircapsid protein thereby maintaining the genotype–phenotype link-age necessary for directed evolution. Virus production was deter-mined by real-time PCR. Transfection with 600 ng of libraryplasmid was the lowest concentration that yielded viruses.

Forty-eight hours after transfection with 600 ng of library plas-mid, viruses were harvested, treated with benzonase and purifiedvia CsCl centrifugation. After isolation of peak viral fractions, thesingle stranded DNA genome was amplified, cloned and individualtransformants sequenced. Not surprisingly, the most frequentlyidentified sequence was wild-type, representing 13 out of 20clones sequenced. These were members of the library and not clon-

targeted region, with more deletions (filled arrows) than duplications (open arrows)e first nucleotide of the gene encoding VP1). (B) Examples of the types of mutationslicated sequence.

K. Hida et al. / Archives of Biochemistry and Biophysics 496 (2010) 1–8 5

ing artifacts or contamination because our library capsid gene con-tained a unique silent mutation by design, which was present in all20 clones. Recovery of wild-type capsids validates that our exper-imental procedure can be used to isolate functional capsids from alibrary of mostly non-functional capsids.

Most interestingly, four novel sequences comprised the remain-ing seven clones—two with duplications and two with deletions(Table 1). Unlike typical naïve library members, these library mem-bers were in-frame and all contained short deletions (6 or 9 bp) ortandem duplications (3 or 6 bp).

3.3. Testing of transduction efficiencies in vitro

To compare the ability of our four newly isolated mutants toproduce viral particles compare with wild-type AAV5, recombinantparticles were produced that encoded the GFP-b-actin gene andthen purified by isopycnic CsCl centrifugation. All mutants pro-duced comparable genomic particle titers to the wild-type vectorexcept for R12, which had no detectable viral titer (Table 1). Itwas not apparent whether R12 failed to package genetic materialor whether the particles did not assemble, as unlike AAV2 thereis no capsid ELISA available to look for assembled AAV5 particles.We next tested the infectivity of these viral particles (R9, R31and R39) compared with wild-type. Viral transduction efficiencieswere determined in vitro on two different cell lines: 293T andBEAS-2B (immortalized bronchial epithelial cells). Viruses wereadded to cells at an MOI of 3000 (calculated as DNase I resistantparticles) and after 96 h, expression of the GFP transgene was as-sessed via flow cytometry. R31 and R39 showed some transduction(Table 1). However, R9 showed a moderate improvement in thetransduction of BEAS-2B cells. This was unexpected since selectivepressure was not applied for improved transduction during theprocess for identifying functional capsid variants.

4. Discussion

Recent studies have demonstrated the feasibility of using direc-ted evolution approaches for the creation of large numbers of mu-tants and subsequent screening to identify new AAV vectors.However, the mutations have been either limited to single aminoacid changes dispersed throughout the capsid protein [22,23] ora mixture of different serotype capsids as in the case of gene shuf-fling [35,36]. Here we have demonstrated a new approach to iden-tifying sites throughout the viral capsid that are tolerant todeletions and duplications (i.e. insertions).

Our approach takes advantage of the nature of the ends of dou-ble-stranded DNA cuts produced by DNaseI digestion of plasmidDNA, which upon subsequent blunting of the ends using DNA

Table 1Selected library members and their corresponding particle titers and transduction efficien

Mutationa Occurencesb

Wild-type — 13R9 Dup 574–575 1R12 Dup 492 2R31 D 176–178 1R39 D 492–493 3

a Dup indicates a duplication and D indicates a deletion of the amino acids indicatedb Number of occurrences out of 20.c DNase resistant genomic particle titer (viral particles/ml) of recombinant AAV withd No viral titer was detected for this mutant.e Transduction efficiencies of BEAS-2B and 293T cells, expressed as number of transduc

fluorescence was quantified 96 h later by FACS. Transduced cells were defined as cellindependent experiments with each condition done in triplicates with >5000 cells coun

f No transduction was detected.

polymerase creates deletions and tandem duplications. We wereable to create a large and diverse library with >1 � 106 transfor-mants after step 4 in the library construction scheme (Fig. 1A)and �1 � 105 transformants after each subsequent transformationstep (steps 7 and 10). Mutations of variable lengths were distrib-uted throughout our target gene, with the majority of the muta-tions (deletions or duplications) being relatively small in length.Except for the largest deletion (119 bp), all mutations ranged from4 to 27 bases, which would correspond to roughly 1–9 amino acids,a size comparable to that of foreign peptides inserted into AAVcapsids in previous studies [14–16].

Due to the inherent nature of the mutations we made, fewviruses were expected to be still infectious. We identified four un-ique mutants after selecting for packaging in 293T cells andsequencing 20 clones. Sequencing of more clones will likely resultin the identification of additional functional sequences; however,since two clones were found more than once, this suggests thatthe number of additional functional sequences may not be exten-sive. The sequences identified all have small changes (1–3 aminoacids) relative to those found in a random sequencing of the naïvelibrary. Even with the use of the ‘marker’ sequence, wild-type con-stituted almost 65% of isolated virions. These sequences presum-ably resulted from DNAse I digestions that produced blunt endsand are unavoidable. If precautions had not been taken to mini-mize wild-type sequences in our final library, it is unlikely thatwe would have been able to isolate any mutants as they wouldhave been overwhelmed by the number of wild-type sequences.Thus, minimization of wild-type sequences using the marker se-quence was critical in the identification of new capsid sequences.A selection scheme that does not favor wild-type AAV (i.e. theuse of cells that are poorly infected by AAV5) will also likely reducethe number of wild-type sequences identified as well as favor no-vel sequences with improved transduction properties on such cells.

Although this library was subjected to selection for the ability toassemble into viruses and not for any improved properties it isencouraging that we were able to identify a clone with improvedtransduction abilities. The two most commonly identified muta-tions, R12 and R39 occurs at position 492, possibly suggesting thatthis site is highly tolerant to changes. However, recombinant GFP-R39 had negligible transduction of both 293T and BEAS-2B cellsand GFP-R12 did not produce a detectable viral titer. This may indi-cate that mutations near this area may interrupt a critical capsidinteraction necessary for both transduction and particle assemblyor genome packaging.

The sequence alignment of five different AAV serotypes (1, 2, 4,5, and 6) shows that three of the four mutations identified here(R12, R31, R39) occur in regions of low sequence conservation(Fig. 3). The R31 deletion continues a large deletion that occurs

cies of recombinant AAV-GFP-actin.

rAAV-GFP titerc Transduction efficiencye

BEAS-2B 293T

3.9E10 1200 ± 120 200 ± 182.8E10 3400 ± 340 240 ± 10—d N/A N/A3.2E10 245 ± 50 30 ± 203.5E10 50 ± 12 —f

by the numbers. Numbering is based on the amino acid sequence of VP1.

GFP-actin was assessed by real-time PCR.

ed cells/108 viral particles. Viruses were added to cells at a genomic MOI of 3000 ands with fluorescence greater than 99% of untransduced cells. Data represents n = 3ted. N/A indicates that the experiment was not performed.

6 K. Hida et al. / Archives of Biochemistry and Biophysics 496 (2010) 1–8

in the AAV5 sequence. In contrast, the most promising mutant, R9,contains an insertion between glutamine and serine, two residuesconserved in at least four of the five serotypes. This site is also oneamino acid away from a site previously shown to be amenable topeptide insertion on the AAV2 surface [14]. This supports the ideathat this method can successfully identify sites that are tolerant tolarger peptide insertions. Although the R9 mutation lies close tothe heparin binding site, the primary receptor interaction for

Fig. 3. Sequence alignment of Cap genes of AAV 1, 2, 4, 5, and 6. Alignments were madeare highlighted in black. Sites within the AAV2 capsid previously shown to be amenable tblue. All insertions occur immediately after the colored residue and deletions are colorereferred to the web version of the article.)

AAV2, AAV5 does not bind to heparin sulfates at this location[29]. Although atomic resolution models of AAV5 are not currentlyavailable, by comparing the sequence alignment with AAV2 andsuperimposing onto the solved structure of AAV2, it appears thatR9, R12, and R39 are all on the surface (Fig. 4). Both locations ap-pear close to sites other researchers have identified in AAV2 thatare tolerant to insertions (red) [13–16]. The D 176–178 deletionof R31 is not depicted in this figure as it lies outside the solved

using the MACAW software. Residues that are highly conserved across all serotypeso insertions [13–16] are shown in red, and sites in AAV5 identified here are shown ind in grey. (For interpretation of color mentioned in this figure legend the reader is

Fig. 4. Molecular model of the AAV2 capsid trimer. The surface of the trimeric unit of AAV2 is highlighted from a top view and from a side view. Locations of the mutationsidentified on the AAV5 surface (D 492–493, Dup 492 and Dup 574) are approximated by sequence alignment with the AAV2 genome and are shown in blue and green.Identified mutations are exaggerated in size for clarity. Previously identified sites in AAV2 that are tolerant to insertions [13–16] are shown in red. (For interpretation of colormentioned in this figure legend the reader is referred to the web version of the article.)

K. Hida et al. / Archives of Biochemistry and Biophysics 496 (2010) 1–8 7

region within AAV2. These exact mutations have not been previ-ously identified; however D 492–493 and Dup 492, occur in a re-gion with low sequence homology, such that it is difficult topinpoint exactly where on the AAV2 structure this mutation inAAV5 corresponds.

Although the mutations that led to a functional virus were allsmall mutations, it is reasonable to expect that these sites maybe tolerant to insertions of larger peptides. Large tandem duplica-tions may be more difficult to produce, as evidenced by the pre-dominance of deletions in our naïve library. Further screening ofthis library with a selective pressure that does not favor wild-typemay identify novel mutants and/or identify larger scale changes inthe capsid. Although R9, the most successful mutant identifiedhere, lies close to a previously identified site on AAV2 and mayhave been identifiable via rational design, the random mutagenesismethod introduced here should be broadly applicable to any virus,including those with unsolved structures not amenable to rationaldesign approaches. Viruses more tolerant to changes in the capsidmay yield more variants, leading to further understanding of theviral surface for future engineering.

Acknowledgments

This research was supported in part by the Intramural ResearchProgram of the NIH and NIDCR to J.A.C.

References

[1] C.K. Conrad, S.S. Allen, S.A. Afione, T.C. Reynolds, S.E. Beck, M. Fee-Maki, X.Barrazza-Ortiz, R. Adams, F.B. Askin, B.J. Carter, W.B. Guggino, T.R. Flotte, GeneTher. 3 (1996) 658–668.

[2] C.L. Halbert, E.A. Rutledge, J.M. Allen, D.W. Russell, A.D. Miller, J. Virol. 74(2000) 1524–1532.

[3] S.G. Sumner-Jones, L.A. Davies, A. Varathalingam, D.R. Gill, S.C. Hyde, GeneTher. 13 (2006) 1703–1713.

[4] X. Xiao, J. Li, R.J. Samulski, J. Virol. 70 (1996) 8098–8108.[5] G.P. Niemeyer, R.W. Herzog, J. Mount, V.R. Arruda, D.M. Tillson, J. Hathcock,

F.W. van Ginkel, K.A. High, C.D. Lothrop Jr., Blood 113 (2009) 797–806.[6] S. Daya, K.I. Berns, Clin. Microbiol. Rev. 21 (2008) 583–593.

[7] C. Mueller, T.R. Flotte, Gene Ther. 15 (2008) 858–863.[8] J.B. Nguyen, R. Sanchez-Pernaute, J. Cunningham, K.S. Bankiewicz, Neuroreport

12 (2001) 1961–1964.[9] D. Duan, Y. Yue, Z. Yan, P.B. McCray Jr., J.F. Engelhardt, Hum. Gene Ther. 9

(1998) 2761–2776.[10] R.W. Walters, T. Grunst, J.M. Bergelson, R.W. Finberg, M.J. Welsh, J. Zabner, J.

Biol. Chem. 274 (1999) 10219–10226.[11] G. Wang, B.L. Davidson, P. Melchert, V.A. Slepushkin, H.H. van Es, M. Bodner,

D.J. Jolly, P.B. McCray Jr., J. Virol. 72 (1998) 9818–9826.[12] N. Chirmule, K. Propert, S. Magosin, Y. Qian, R. Qian, J. Wilson, Gene Ther. 6

(1999) 1574–1583.[13] A. Girod, M. Ried, C. Wobus, H. Lahm, K. Leike, J. Kleinschmidt, G. Deleage, M.

Hallek, Nat. Med. 5 (1999) 1052–1056.[14] W. Shi, G.S. Arnold, J.S. Bartlett, Hum. Gene Ther. 12 (2001) 1697–1711.[15] L. Perabo, H. Buning, D.M. Kofler, M.U. Ried, A. Girod, C.M. Wendtner, J. Enssle,

M. Hallek, Mol. Ther. 8 (2003) 151–157.[16] P. Wu, W. Xiao, T. Conlon, J. Hughes, M. Agbandje-McKenna, T. Ferkol, T. Flotte,

N. Muzyczka, J. Virol. 74 (2000) 8635–8647.[17] J.E. Rabinowitz, W. Xiao, R.J. Samulski, Virology 265 (1999) 274–285.[18] N.A. Huttner, A. Girod, L. Perabo, D. Edbauer, J.A. Kleinschmidt, H. Buning, M.

Hallek, Gene Ther. 10 (2003) 2139–2147.[19] C.E. Wobus, B. Hugle-Dorr, A. Girod, G. Petersen, M. Hallek, J.A. Kleinschmidt, J.

Virol. 74 (2000) 9281–9293.[20] Q. Xie, W. Bu, S. Bhatia, J. Hare, T. Somasundaram, A. Azzi, M.S. Chapman, Proc.

Natl. Acad. Sci. USA 99 (2002) 10405–10410.[21] Z. Wu, A. Asokan, J.C. Grieger, L. Govindasamy, M. Agbandje-McKenna, R.J.

Samulski, J. Virol. 80 (2006) 11393–11397.[22] N. Maheshri, J.T. Koerber, B.K. Kaspar, D.V. Schaffer, Nat. Biotechnol. 24 (2006)

198–204.[23] L. Perabo, J. Endell, S. King, K. Lux, D. Goldnau, M. Hallek, H. Buning, J. Gene

Med. 8 (2006) 155–162.[24] M.J. Blankinship, P. Gregorevic, J.M. Allen, S.Q. Harper, H. Harper, C.L. Halbert,

D.A. Miller, J.S. Chamberlain, Mol. Ther. 10 (2004) 671–678.[25] Y. Liu, T. Okada, K. Sheykholeslami, K. Shimazaki, T. Nomoto, S. Muramatsu, T.

Kanazawa, K. Takeuchi, R. Ajalli, H. Mizukami, A. Kume, K. Ichimura, K. Ozawa,Mol. Ther. 12 (2005) 725–733.

[26] F. Mingozzi, J. Schuttrumpf, V.R. Arruda, Y. Liu, Y.L. Liu, K.A. High, W. Xiao, R.W.Herzog, J. Virol. 76 (2002) 10497–10502.

[27] B. De, A. Heguy, P.L. Leopold, N. Wasif, R.J. Korst, N.R. Hackett, R.G. Crystal, Mol.Ther. 10 (2004) 1003–1010.

[28] J. Zabner, M. Seiler, R. Walters, R.M. Kotin, W. Fulgeras, B.L. Davidson, J.A.Chiorini, J. Virol. 74 (2000) 3852–3858.

[29] G. Di Pasquale, B.L. Davidson, C.S. Stein, I. Martins, D. Scudiero, A. Monks, J.A.Chiorini, Nat. Med. 9 (2003) 1306–1312.

[30] U. Bantel-Schaal, H. Delius, R. Schmidt, H. zur Hausen, J. Virol. 73 (1999) 939–947.[31] J.A. Chiorini, F. Kim, L. Yang, R.M. Kotin, J. Virol. 73 (1999) 1309–1319.[32] M. Ostermeier, A.E. Nixon, J.H. Shim, S.J. Benkovic, Proc. Natl. Acad. Sci. USA 96

(1999) 3562–3567.

8 K. Hida et al. / Archives of Biochemistry and Biophysics 496 (2010) 1–8

[33] G. Di Pasquale, A. Rzadzinska, M.E. Schneider, I. Bossis, J.A. Chiorini, B. Chiorini,Mol. Ther. 11 (2005) 849–855.

[34] G. Guntas, M. Ostermeier, J. Mol. Biol. 336 (2004) 263–273.[35] J.T. Koerber, J.H. Jang, D.V. Schaffer, Mol. Ther. 16 (2008) 1703–1709.

[36] W. Li, A. Asokan, Z. Wu, T. Van Dyke, N. DiPrimio, J.S. Johnson, L.Govindaswamy, M. Agbandje-McKenna, S. Leichtle, D.E. Redmond Jr., T.J.McCown, K.B. Petermann, N.E. Sharpless, R.J. Samulski, Mol. Ther. 16 (2008)1252–1260.